Skip to content

AI Rap Flow GUIDE

How to Do the Outstanding Tiny Desk Trend

A practical five-photo guide to the fixed AI performance.

Short answer

Choose a photo for each of five fixed roles, submit the Outstanding template, then review and download the 15-second result. The preview and downloaded MP4 include audio. All five performer slots must be filled before Generate is available. Open the Outstanding Tiny Desk studio.

A current AI-generated example made with the five-role workflow.

What is the Outstanding Tiny Desk trend?

The Outstanding Tiny Desk trend is a fixed-format AI music performance built around five named performer positions and a short sequence of camera cuts. Instead of asking you to write a song or invent a stage, the workflow starts with five photos and uses them as visual references for a prepared performance. It is intended for someone who wants to see a small band setup, a red keyboard, a featured singer and a saxophone performer in one compact vertical video.

The roles stay attached to their slots: Lead, Keys, Backup L, Backup R and Sax. Each role needs its own photo. The five portraits do not become interchangeable once generation begins, so role assignment is one of the most important choices you make. The Lead is the featured singer. Keys is the keyboard player. Backup L and Backup R are supporting performers positioned on opposite sides. Sax is the horn player featured in the final close-up.

The product creates a new video from those references and a fixed motion template. It does not promise a frame-for-frame recreation of any source clip. Faces, gestures, timing and framing can vary between generations. Think of the result as a new interpretation of the arrangement, not an editing tool that pastes people into an existing recording.

What does an Outstanding Tiny Desk AI video include?

The output is a 15-second vertical video in a 9:16 frame. The motion follows three parts: the keyboard trio opens the performance, the camera cuts to the Lead, and the final close-up features Sax. The reference timings are 0–5.25 seconds for the group at the keyboard, 5.25–11.54 seconds for the Lead, and 11.54–14.15 seconds for Sax. The remaining time completes the final shot; the sequence does not return to a fourth group shot.

The order is fixed, so you cannot move the saxophone shot to the opening or change the number of cuts. A fixed template makes the workflow simpler: you supply the cast, choose an available output quality, review the request and submit it. It also means the output is less flexible than a custom music video generator. There are no lyric, song, role-order or duration controls in this trend.

Sound is part of the Outstanding example and generated output. The preview and downloaded MP4 include audio. That audio setup is fixed along with the visual sequence, so this is not the place to upload a replacement song or ask for a different genre. If you want to create a new rap or pop song and edit its words, the separate Tiny Desk AI Rap & Pop workflow is a better fit.

How to make an Outstanding Tiny Desk AI video

1. Prepare five separate portraits

Collect one photo for each role before opening the generator. Use images you have permission to use, and make sure the person in each photo is an adult. A group photograph is not a substitute for five separate uploads because the tool maps each file to a specific performer slot.

For a useful portrait, choose a clear image with the face visible, enough light to distinguish the eyes and hair, and a crop that includes the head and shoulders. Avoid heavy filters, motion blur, strong shadows across the face and large objects covering the person. The generated scene has its own lighting and camera movement, so a clean reference gives the model a clearer visual guide.

The source photos do not need to look like a finished music video frame. They should show the person clearly. If you are choosing between a dramatic image and a simple, well-lit portrait, the simple image is often easier to map consistently to a fixed role.

2. Match each photo to its role

Upload the featured vocalist into Lead, the keyboard performer into Keys, the left backing performer into Backup L, the right backing performer into Backup R, and the horn performer into Sax. Confirm every label before continuing. If two people are swapped, the generated result may still be visually coherent, but the intended cast order will be wrong.

Use the on-screen role labels as the source of truth. Do not infer a role from upload order after rearranging files on your computer. A quick check is to read the five slots from top to bottom and say the assignment out loud: Lead, Keys, Backup L, Backup R, Sax. The generator keeps the fixed role mapping when it builds the request.

3. Choose quality and check the credit cost

Select one of the quality options currently shown in the studio. The available options and credit cost are displayed before you submit. Check both before generation, especially if you are deciding between a lower and higher output resolution. This trend currently uses the same short vertical composition at either available quality; quality changes the delivered resolution, not the story order.

If the credit balance is too low, the app may offer packs or plans. Review the product before leaving the generator. A purchase is a separate checkout step and does not submit the video task automatically. Once credits are available, return to the studio and review the task before you generate.

4. Review the request and submit once

Before creating, check the five assigned photos, quality and displayed credit amount. Read the permission prompt and submit only if you have the right to use each image. The generation request reserves the shown credits. Avoid opening several copies of the same submission or repeatedly clicking a retry control while the service is still checking the first request.

The page shows a task reference after a submission is accepted. Use the task status and your My videos library to follow progress. A task reference means the request was recorded; it does not mean the video has finished rendering. Wait for a successful status before downloading.

5. Preview the result before sharing

Watch the complete video, not just the opening frame. Check that each person appears in the intended part of the performance, that the result has sound, and that the last shot has completed. AI-generated faces and movement can differ from the input portraits. If the result has an unexpected expression or framing, decide whether it is still suitable before sharing it publicly.

Download the completed MP4 from the task result or My videos. Keep a local copy if you plan to use the video later. Preview and download refer to the generated result; the browser preview is not a guarantee that every social platform will display the file in the same way.

Tips for five consistent performer references

Use one clearly identifiable subject per photo. When a portrait contains multiple people, the model may have less certainty about which person should guide the role. Keep the face large enough to see and choose a pose that fits the assigned performer. The Keys and Sax references are especially helpful when the instrument or seated position is visible, although the generated scene still follows the fixed template.

Keep the five images reasonably consistent in sharpness and lighting. The pictures do not need identical backgrounds, but extreme differences—one very dark image beside four bright close-ups—can create uneven results. A neutral background is fine. The generated stage supplies the setting; your primary job is to provide clear identity references.

Do not use a portrait simply because it matches a celebrity or a performer from a reference video. Use photos you own or have permission to use, and do not imply that a person endorsed the result. If the video is for a public post, confirm that the people pictured are comfortable with the intended use. Those checks remain your responsibility even when an AI tool creates the final video.

Fixed trend or original Tiny Desk AI rap?

Choose Outstanding when you want a five-person cast, a fixed camera sequence and a short performance with audio already included. It works best when you have five photos ready and want the template to handle the staging. You do not need to write lyrics or compose a song for this path.

Choose Tiny Desk AI Rap & Pop when you want to shape an original song. That workflow starts with a Lead photo and lets you add an optional Partner, choose RAP or POP, provide a topic and edit lyrics or a hook. It does not require separate Keys, Backup L, Backup R and Sax portraits. Its controls and output are different from the fixed Outstanding template.

If you are unsure, decide what you want to control. If the fixed band arrangement is the point, use Outstanding. If the words and musical direction matter more, use the original rap and pop studio. The Tiny Desk AI overview compares both paths in more detail.

Troubleshooting common Outstanding Tiny Desk results

If a performer appears in the wrong role, first confirm that each portrait was uploaded to its matching label. The template uses the assigned photo for that role; it does not infer a role from a person’s clothing or instrument. If the mapping was correct but the generated face still varies, remember that references guide a result rather than lock every frame to a source image.

If the clip does not play sound in a browser, check the player volume and then download the MP4 to test it locally. The current example includes an audio track. Browser autoplay policies may keep video muted until you press play, and a muted preview is not evidence that the downloaded MP4 is silent.

If you need a different song, duration or camera order, this is a workflow mismatch rather than a missing setting. Use Tiny Desk AI Rap & Pop for editable lyrics and a new song, or choose another trend whose controls fit the idea. Avoid submitting the same job repeatedly while a task is still processing; follow its status in My videos.

Outstanding Tiny Desk FAQ

How many photos does the Outstanding Tiny Desk generator need?

It needs five separate photos: one each for Lead, Keys, Backup L, Backup R and Sax. Every slot must be filled before you can submit.

Can I change the camera sequence or video length?

No. The template uses a fixed 15-second vertical sequence with the group, Lead and Sax shots in that order. It does not provide duration or shot-order controls.

Does the Outstanding Tiny Desk preview have sound?

Yes. The preview and downloaded MP4 include audio. The audio selection is fixed for this template.

Will my result look exactly like the example?

No. The photos guide the performer identities, and the template guides the staging, but AI-generated faces, movement, timing and framing can vary. Review the finished clip before posting.

Can I write lyrics or upload a different song?

No. Outstanding is a fixed performance workflow. For editable words and a newly generated rap or pop song, use Tiny Desk AI Rap & Pop.

Can I use any portrait I find online?

Use photos you own or have permission to use. Make sure everyone pictured is an adult and is comfortable with the intended use of the generated result.

Read the Tiny Desk AI format guide, open the Outstanding Tiny Desk studio, or browse all music trends.